The recent development of microarray platforms, capable of genotyping hundreds of thousands of single nucleotide polymorphisms (SNPs), has resulted in the rapid identification of novel susceptibility genes for a number of complex diseases. However, recent data from several large-scale genome-wide association studies (GWAS) of schizophrenia (SCZ) have been disappointing; it is clear that identification of even several common variants of very modest effect will leave considerable genetic variance in SCZ unexplained. In this EUREKA application, we challenge the implicit assumptions underlying GWAS: that multiple common alleles of additive effect combine to produce the phenotype, and suggest a complementary hypothesis: that a proportion of cases of schizophrenia may be characterized by a simpler genetic architecture, and that one or more homogeneous molecular subtypes may co-exist alongside the polygenic pool of cases. Support for this model comes from a novel statistical approach to analyzing GWAS data, which we have termed "whole genome homozygosity analysis" (WGHA). Applied to our own SCZ dataset (PI: Anil Malhotra), WGHA has identified several rare recessive loci of high penetrance, encompassing several chromosomal regions (of approximately 200kb-2MB) that have been identified in prior SCZ linkage and association studies. This EUREKA proposal aims to extend this initial finding in several stages. First, we will apply WGHA analysis to a much larger SCZ dataset derived from the Genetic Analysis Information Network (GAIN) initiative. Next, we will re-analyze data from a whole genome linkage dataset (PI: Hugh Gurling), results of which will guide a novel SNP-based linkage study of SCZ pedigrees from the NIMH repository. Analysis of linkage data will be focused on identifying small clusters of pedigrees with strongly recessive transmission of similar loci. Highly penetrant recessive loci identified and replicated across these analyses will then be interrogated in both case and control samples using next-generation high-throughput deep resequencing technology (Illumina 1G platform), in collaboration with Cold Spring Harbor Laboratory (PI: W. Richard McCombie). By the end of the first half of the proposed 3-year funding period, we aim to have characterized one or more recessive subtypes of schizophrenia, accounting for up to 10% of all cases. At the completion of the project in Year 3, we aim to have identified, through deep resequencing, one or more highly penetrant mutations underlying these cases. Schizophrenia (SCZ) constitutes the fifth leading cause of disability in the US. We aim to use novel statistical approaches and state-of-the-art genomic sequencing technology to characterize one or more homogeneous genetic subtypes of schizophrenia, and to identify the causal mutation(s) underlying such subtype(s). Findings will create new opportunities for diagnosis and prediction of schizophrenia, and for understanding its biology. [unreadable] [unreadable] [unreadable] [unreadable]